The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Deverberasi pertuturan adalah salah satu tugas paling sukar dalam pemprosesan isyarat akustik. Daripada pelbagai masalah yang terlibat dalam tugasan ini, kertas kerja ini menyerlahkan "pemutihan berlebihan," yang meratakan ciri-ciri pertuturan pulih. Herotan ini kadangkala berlaku apabila penapis songsang dikira terus daripada isyarat mikrofon. Kertas kerja ini meninjau dua kajian yang berkaitan dengan masalah ini. Kajian pertama menunjukkan kemungkinan mengimbangi pemutihan berlebihan sedemikian untuk mencapai pertuturan-dereverberation yang tepat. Kajian kedua membentangkan pendekatan baharu untuk menghampiri ucapan asal dengan membuang kesan pantulan lewat daripada ucapan bergema yang diperhatikan.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Masato MIYOSHI, Marc DELCROIX, Keisuke KINOSHITA, "Calculating Inverse Filters for Speech Dereverberation" in IEICE TRANSACTIONS on Fundamentals,
vol. E91-A, no. 6, pp. 1303-1309, June 2008, doi: 10.1093/ietfec/e91-a.6.1303.
Abstract: Speech dereverberation is one of the most difficult tasks in acoustic signal processing. Of the various problems involved in this task, this paper highlights "over-whitening," which flattens the characteristics of recovered speech. This distortion sometimes happens when inverse filters are directly calculated from microphone signals. This paper reviews two studies related to this problem. The first study shows the possibility of compensating for such over-whitening to achieve precise speech-dereverberation. The second study presents a new approach for approximating the original speech by removing the effect of late reflections from observed reverberant speech.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e91-a.6.1303/_p
Salinan
@ARTICLE{e91-a_6_1303,
author={Masato MIYOSHI, Marc DELCROIX, Keisuke KINOSHITA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Calculating Inverse Filters for Speech Dereverberation},
year={2008},
volume={E91-A},
number={6},
pages={1303-1309},
abstract={Speech dereverberation is one of the most difficult tasks in acoustic signal processing. Of the various problems involved in this task, this paper highlights "over-whitening," which flattens the characteristics of recovered speech. This distortion sometimes happens when inverse filters are directly calculated from microphone signals. This paper reviews two studies related to this problem. The first study shows the possibility of compensating for such over-whitening to achieve precise speech-dereverberation. The second study presents a new approach for approximating the original speech by removing the effect of late reflections from observed reverberant speech.},
keywords={},
doi={10.1093/ietfec/e91-a.6.1303},
ISSN={1745-1337},
month={June},}
Salinan
TY - JOUR
TI - Calculating Inverse Filters for Speech Dereverberation
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1303
EP - 1309
AU - Masato MIYOSHI
AU - Marc DELCROIX
AU - Keisuke KINOSHITA
PY - 2008
DO - 10.1093/ietfec/e91-a.6.1303
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E91-A
IS - 6
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - June 2008
AB - Speech dereverberation is one of the most difficult tasks in acoustic signal processing. Of the various problems involved in this task, this paper highlights "over-whitening," which flattens the characteristics of recovered speech. This distortion sometimes happens when inverse filters are directly calculated from microphone signals. This paper reviews two studies related to this problem. The first study shows the possibility of compensating for such over-whitening to achieve precise speech-dereverberation. The second study presents a new approach for approximating the original speech by removing the effect of late reflections from observed reverberant speech.
ER -